177 shares
Toyota driver Kyle Busch (18) battles with Ford drivers Brad Keselowski (2) and Joey Logano (22) as well as Chevrolet driver Martin Truex Jr. (78) at the NASCAR Sprint Cup race at Kentucky Speedway. (Christopher Hanewinckel, USA TODAY Sports)
LOUDON, N.H. — Somewhere in climate-controlled rooms in suburban Charlotte industrial parks, three sets of super computers churned through innumerable data points and calculations on Friday.
As Chevrolet, Ford and Toyota Sprint Cup teams practiced at New Hampshire Motor Speedway, their engineers back in NASCAR’s hub wielded computing power upon upcoming races with new rules packages at Indianapolis Motor Speedway, Michigan International Speedway, and Darlington Raceway.
Although motorsports is ever-technical and engineering-driven, the mastery of aerodynamics that determine the performance of race cars and, somewhat frustratingly for the engineers tasked with it, the quality of the show, remains a tricky proposition.
The computers keep working on it.
PREVIEW: Five storylines to watch at New Hampshire Sprint Cup race
NASCAR’s decision, announced to teams a month before implementation, to deploy track-specific aero packages for Kentucky Speedway last week, Indianapolis next week, Michigan and Darlington underscored the fact that perceived entertainment value is a paramount concern. Everyone understands, said General Motors Racing director Mark Kent, whose program has won 12 consecutive manufacturers’ titles. And so they, and their computers, adjust and keep grinding.
“Corporately, we totally understand the value of participating in motorsports,” Kent told USA TODAY Sports. “It’s as important to us as it is the sanctioning body to assure that the racing is exciting and can collectively maintain that fan base so we can show our products off in front of them. It’s as important to us and it is to the sanctioning body. And I would say it’s important to us try and stay on top as rules change.”
Kent has had to address the equation in both NASCAR and IndyCar. In the open wheel series in June, a pitched race at Auto Club Speedway in Fontana, Calif., featured high-speed, close-proximity racing, and a rollover accident at the finish titillated fans and divided drivers over the aerodynamic thrill ride that is pack racing.
“It is very subjective,” Kent said of perceptions of what makes a race entertaining. “It is not always clear what defines good racing. There’s people that love pack racing. There’s people that don’t like pack racing. There’s people that like the slingshot effect, that don’t like that. Everybody has their own opinion. I think that’s the biggest challenge to all these sanctioning bodies, trying to understand what the fans feel is important to make the racing exciting.”
And that can be exasperating to engineers trained to dissect variables and hone performance, whose computation fluid dynamics models, simulators and transient analyses don’t deduce entertainment value. But they reveal the troubling variable.
“It does get pretty frustrating, and it all comes down to money, really,” said Toyota Racing Development group vice president and technical director Andy Graves. “In the last 20 years, as the sport grew, and so much more money was in the sport, in order to vie for the best sponsors and to get the best drivers all the teams invested in themselves and invested in this technology, and the more science that we have and we push every part of the car to the extremes, in essence you’re getting a good lap time out of that. But you have to have that in the perfect world. And when all of a sudden you don’t have the perfect world anymore, i.e. you’re stuck in traffic, you’re not going to be as efficient and you’re not going to have the performance as when you were in clean air.”
Much of the difficulty in designing race cars that produce desirable race qualities — especially the ability to maneuver in close quarters and pass freely — is honing their stability in actual race conditions. Super computers excel in developing swift and stable cars when racing alone. Modeling remains difficult, however, in producing a car that handles anywhere remotely as well in turbulent or so-called “dirty” air that swirls off the curves and nuances of race cars ahead and alongside.
“That is the very difficult part because the air can come off, in our series, with different cars, can come off Ford, Chevy and Toyota a little differently,” Ford Performance Motorsports Engineering manager Mark Rushbrook told USA TODAY Sports. “So when you’re the trailing car, you’ve got to deal with that, and then depending on where you are, how close you are to the car that’s in front of you to the right a little bit, to the left a little bit, whether you get alongside it.
“That’s where, as good as out tools are, that’s where the limitation really is. And a lot of it is the amount of CPU to make all those runs to do a full analysis for the trailing car.”
Reducing downforce made Sprint Cup cars less-perfect engineering-wise and more difficult to drive at Kentucky, but produced a race that was universally lauded by drivers, fans and the series. That set-up will be utilized at Darlington while higher-drag packages are planned at Indianapolis and Michigan. And after a three-day test at Chicagoland Speedway, more changes are possible.
Aerodynamics has also been a weekly topic in the IndyCar series with the first implementation of body kits developed by manufacturers Chevrolet and Honda. Chevrolets have won nine of 12 races heading into the event at Iowa Speedway Saturday night, but the manufacturer has had problems to address, too. Three airborne flips by Chevrolets in pre-Indianapolis 500 practice forced the series to mandate safety modifications. Honda teams have become more proficient in manipulating the myriad of adjustable winglets in an attempt to keep up with its rival but remain behind.
The IndyCar race at Fontana featured high-speed, close-proximity racing that thrilled most fans but frustrated some drivers. (Gary A. Vasquez, USA TODAY Sports)
Computers will have much to do with catching up. Technology has become the primary method of developing cars with teams prohibited by NASCAR and INDYCAR rules from conducting private on-track tests. Graves said that feedback from drivers in cars “definitely has an impact, for sure” in the development. But advances in simulation seem to have appeased both engineers and drivers that quality cars can be devised through computation as well as endless hours circling race tracks.
“(Driver input) is valuable because not all drivers are created equal,” Kent concurred. “Although on paper a specific configuration may be the quickest, it may not be the quickest for a particular driver. So the driver input is important and that is why you’re seeing more reliance going forward on driving simulators, driver models and track simulations. We’re trying to either mathematically model or get driver influence through other means other than being actually in a race car, on a track, burning fuel and using tires.”
INDYCAR: Can Andretti Autosport turn season around at Iowa?
Driver sentiment tends to support the process, partly because they understand the forces of cost-reduction and technology-inclusion are powerful. Some key traffic situations are nearly impossible to replicate on track, anyway, Rushbrook said. Driver Joey Logano said open testing “doesn’t seem like it needs to come back” even though computer modeling also has some fallibility.
“All the teams have developed other ways to test,” Logano told USA TODAY Sports. “We’ve gotten pretty strong at simulation, whether it’s seven-post or K rigs or computer simulations. That has developed so much because we can’t go to the racetrack. And it’s cheaper than going to the racetrack. You can do this throughout the whole week.”
Each manufacturer provides assistance in computer modeling and in some cases in procuring wind tunnel time for their teams, who then apply their own engineers and simulations to hone individual results. Graves said a “quality” race engine requires an expenditure of “several million dollars per car” and estimated that teams spend roughly half that amount on aerodynamics.
According to Kent, each CFD test case requires nearly 20 hours on more than 1,000 computer processors working simultaneously. Honda Performance Development president Art St. Cyr, in decrying being subjected to the same pre-Indianapolis 500-qualifying adjustments following Chevrolet’s three flips, said Honda had consumed “2,059,200 CPU hours” alone in analyzing car stability.
Computer modeling undoubtedly helped Sprint Cup teams — especially Toyota-driving Joe Gibbs Racing, which put four drivers in the top five — adjust quickly to changes at Kentucky.
“We put a huge amount of effort into that race, even to the point where we had to question ourselves if we were putting too much time into this new package where it’s one race or two races,” Graves said. “At that time we only knew it was going to be Kentucky and Darlington. The question was whether it was taking our focus off the current package too much. Obviously, when it’s all said and done and we end up with four cars in the top five, we all felt pretty good about the amount of effort we put into it.”
The computers didn’t feel anything. They just kept working.
chevrolet, Ford Motor, honda, indycar, iowa, Joey Logano, kentucky, NASCAR, new hampshire, Toyota, Indycar
177 shares